RankingKimi-K2.6

Kimi-K2.6

Compare models

Data updated 22 Nov 2026

Kimi-K2.6 by Moonshot is a Kimi model. This profile collects the published benchmark results and reported effort settings for this model. Configurations are compared on the effort page. Ranks are on the leaderboard.

69 published benchmark measures · 46 measured effort results · 17 benchmark families contribute across 7 task areas. 4 capability estimates available in the full profile. See all results ↓ · Compare published benchmarks →

Compare with

Choose a second model and its effort to see both profiles and benchmark differences below.

Compare with

Choose a configuration to compare. The overview below combines evidence across settings and has no model-wide rank.

Model evidence summary

This profile combines published settings. It is not a runnable configuration or a leaderboard rank. Compare measured configurations →

Performance profile

Capabilities

Adjusted comparison score · 0–100
Kimi-K2.6 capability radarComparison scores from 0 to 100. Unknown capabilities have no point; lines stop at gaps. Hollow points are preliminary. Exact values and support labels follow the chart.AgenticHard reasoningCodingHuman prefKnowledgeMultimodalLong context50100Kimi-K2.6 · Agentic: 12.8 · SupportedKimi-K2.6 · Hard reasoning: 33.6 · SupportedKimi-K2.6 · Coding: 8.6 · SupportedKimi-K2.6 · Multimodal: 12.9 · Preliminary

Filled points are supported; hollow points are preliminary. Lines stop at unknown capabilities. Exact values and sources follow below.

Agentic4 families · 0 with independent evidence · Supported12.8

4 core families; 9 direct opponents across 7 labs.

Estimated against the same complete panel: gpt-5.6-sol claude-opus-4-8 gemini-3.1-pro . Indirect comparisons assume performance can be summarized by one strength within this capability.

Refit without one family: 9.1–15.8; 0/13 scenarios unsupported. Without one publisher: 11.1–95.4; 0/15 unsupported. Smoothing check: 9.4–20.6; 0/3 unsupported. These are sensitivity checks, not confidence intervals.

Audit: family and publisher holdouts improved on neutral predictions within evaluated coverage; source sensitivity still requires inspection. Validation audit.

Matched family win shares below describe source evidence; they are not averaged to produce the adjusted score.

Hard reasoning4 families · 0 with independent evidence · Supported33.6

4 core families; 8 direct opponents across 7 labs.

Estimated against the same complete panel: gpt-5.6-sol claude-opus-4-8 gemini-3.1-pro . Indirect comparisons assume performance can be summarized by one strength within this capability.

Refit without one family: 28.0–42.0; 0/6 scenarios unsupported. Without one publisher: 10.6–74.8; 1/16 unsupported. Smoothing check: 24.6–41.6; 0/3 unsupported. These are sensitivity checks, not confidence intervals.

Audit: family and publisher holdouts improved on neutral predictions within evaluated coverage. Validation audit.

Matched family win shares below describe source evidence; they are not averaged to produce the adjusted score.

Coding4 families · 0 with independent evidence · Supported8.6

4 core families; 12 direct opponents across 8 labs.

Estimated against the same complete panel: gpt-5.6-sol claude-opus-4-8 gemini-3.1-pro . Indirect comparisons assume performance can be summarized by one strength within this capability.

Refit without one family: 5.2–30.2; 0/6 scenarios unsupported. Without one publisher: 5.9–30.3; 0/19 unsupported. Smoothing check: 5.9–15.7; 0/3 unsupported. These are sensitivity checks, not confidence intervals.

Audit: family and publisher holdouts improved on neutral predictions within evaluated coverage. Validation audit.

Matched family win shares below describe source evidence; they are not averaged to produce the adjusted score.

Human pref1 families · 1 with independent evidence · PreliminaryUnknown

1 core families; 41 direct opponents across 12 labs. No connected comparison to the complete reference panel

Estimated against the same complete panel: gpt-5.6-sol claude-opus-4-8 gemini-3.1-pro . Indirect comparisons assume performance can be summarized by one strength within this capability.

Refit without one family: No supported estimate; 1/1 scenarios unsupported. Without one publisher: No supported estimate; 1/1 unsupported. Smoothing check: No supported estimate; 3/3 unsupported. These are sensitivity checks, not confidence intervals.

Audit: cross-family and cross-publisher predictions could not be evaluated from the available comparison network. Validation audit.

Matched family win shares below describe source evidence; they are not averaged to produce the adjusted score.

  • Human preference (Arena): 90.2 observed win share
    lmarena.ai · Source 1
Knowledge1 families · 0 with independent evidence · PreliminaryUnknown

1 core families; 4 direct opponents across 4 labs. No connected comparison to the complete reference panel

Estimated against the same complete panel: gpt-5.6-sol claude-opus-4-8 gemini-3.1-pro . Indirect comparisons assume performance can be summarized by one strength within this capability.

Refit without one family: No supported estimate; 4/4 scenarios unsupported. Without one publisher: No supported estimate; 5/5 unsupported. Smoothing check: No supported estimate; 3/3 unsupported. These are sensitivity checks, not confidence intervals.

Audit: almost no publisher-held-out comparisons remained estimable. Cross-source validity is unresolved. Validation audit.

Matched family win shares below describe source evidence; they are not averaged to produce the adjusted score.

Multimodal2 families · 0 with independent evidence · Preliminary12.9

2 core families; 6 direct opponents across 4 labs. Needs broader benchmark and opponent support

Estimated against the same complete panel: gpt-5.6-sol claude-opus-4-8 gemini-3.1-pro . Indirect comparisons assume performance can be summarized by one strength within this capability.

Refit without one family: No supported estimate; 8/8 scenarios unsupported. Without one publisher: No supported estimate; 8/8 unsupported. Smoothing check: No supported estimate; 3/3 unsupported. These are sensitivity checks, not confidence intervals.

Audit: limited publisher-held-out coverage. Cross-source validity is unresolved. Validation audit.

Matched family win shares below describe source evidence; they are not averaged to produce the adjusted score.

Long context1 families · 0 with independent evidence · PreliminaryUnknown

1 core families; 4 direct opponents across 4 labs. No connected comparison to the complete reference panel

Estimated against the same complete panel: gpt-5.6-sol claude-opus-4-8 gemini-3.1-pro . Indirect comparisons assume performance can be summarized by one strength within this capability.

Refit without one family: No supported estimate; 5/5 scenarios unsupported. Without one publisher: No supported estimate; 5/5 unsupported. Smoothing check: No supported estimate; 3/3 unsupported. These are sensitivity checks, not confidence intervals.

Audit: family and publisher holdouts underperformed neutral predictions. Treat this as an exploratory estimate. Validation audit.

Matched family win shares below describe source evidence; they are not averaged to produce the adjusted score.

  • Aa Lcr: 100.0 observed win share
    NVIDIA · Source 1

Scores estimate outcomes against a shared reference panel; they are not accuracy percentages. Sparse or disconnected evidence cannot qualify an overall profile. Open a capability to inspect its evidence.

Reported effort · Reasoning + unspecified

Settings reported in this model's published benchmark results, including results outside the aggregate. Effort names are provider-specific. These are not API defaults or equal compute budgets.

Mixed settings means multiple settings occur in the evidence. Best across efforts means the source selected its best reported result across settings; it does not mean Max. Unspecified settings stay unknown. This model-summary chart combines reported settings. The leaderboard keeps identified configurations separate and excludes unknown effort.

Inspect each result and its source ↓ · Download effort evidence
Score contributions and missing evidence

17 contributing families across 7 capabilities. Fixed reference panels do not change when the catalog expands.

Capability is fitted jointly across families. Capability estimates below describe different task areas; their weighted sum is not the Capability score.

Results without reviewed compatibility or a reference match remain in the raw evidence below. Coverage counts only contributing results.

Model information & shareable badge
Lab
Moonshot
Catalog status
api-listed
Availability
Documented provider API and open weights for self-hosting
Family
Kimi
Released
—
Context
256,000 tokens
License
—
Model card
https://huggingface.co/moonshotai/Kimi-K2.6
Default Capability family coverage
Documented-evidence family coverage/badge/kimi-k2.6.svg

Kimi-K2.6 measured results by effort setting

Each row is one published result for one effort setting. Evidence collected 2026-11-22. Compare the effort settings →

BenchmarkEffortScoreSource
AA-Briefcase EloReasoning821Artificial Analysis
AA-LCR v1.1Non-reasoning69.7%Artificial Analysis
AA-LCR v1.1Reasoning81.0%Artificial Analysis
Agents' Last ExamHigh9.2%Agents' Last Exam
APEX-Agents-AAReasoning28.5%Artificial Analysis
AutomationBench-AAReasoning13.0%Artificial Analysis
CritPtNon-reasoning1.4%Artificial Analysis
CritPtReasoning8.0%Artificial Analysis
EnterpriseOps-Gym-AAReasoning38.5%Artificial Analysis
Epoch · Chess Puzzles · v2.0.0Unspecified20.0%Epoch AI
Epoch · FrontierMath Tier 4 · v2.0.0Unspecified25.6%Epoch AI
Epoch · FrontierMath Tiers 1–3 · v2.0.0Unspecified57.2%Epoch AI
Epoch · GPQA diamond · v1.0.8Unspecified90.8%Epoch AI
Epoch · Mystery Game Puzzles · v1.0.4 · Tool submissionNon-reasoning12.0%Epoch AI
Epoch · Mystery Game Puzzles · v1.0.4 · Tool submissionUnspecified18.0%Epoch AI
Epoch · OTIS Mock AIME 2024-2025 · v1.1.3Unspecified96.1%Epoch AI
Epoch · SimpleQA Verified · v1.0.0Unspecified34.9%Epoch AI
GDP.pdf: All-pass RateReasoning13.0%Artificial Analysis
GDPval-AA v2.1 LeaderboardReasoning1046Artificial Analysis
GPQA DiamondNon-reasoning78.8%Artificial Analysis
GPQA DiamondReasoning91.1%Artificial Analysis
Humanity's Last ExamNon-reasoning19.6%Artificial Analysis
Humanity's Last ExamReasoning37.5%Artificial Analysis
IFBenchNon-reasoning44.3%Artificial Analysis
IFBenchReasoning76.0%Artificial Analysis
ITBench-AAReasoning31.2%Artificial Analysis
MLCR-AAReasoning11.1%Artificial Analysis
MMMU-ProReasoning79.4%Artificial Analysis
SciCodeReasoning51.5%Artificial Analysis
Terminal-Bench HardNon-reasoning37.9%Artificial Analysis
Terminal-Bench HardReasoning43.9%Artificial Analysis
Terminal-Bench v2.1Reasoning65.9%Artificial Analysis
Terminal-Bench v4.0Reasoning0.5%Artificial Analysis
ThinkingBox-BenchUnspecified37.7%Microsoft and Hugging Face
Toolathlon-VerifiedUnspecified58.0%HKUST NLP
Vals AI · Code Migration v1Unspecified27.8%Vals AI
Vals AI · Excel Modeling Benchmark v1Unspecified57.9%Vals AI
Vals AI · Finance Agent v2Unspecified44.9%Vals AI
Vals AI · Legal Research Bench v1Unspecified15.9%Vals AI
Vals AI · MMLU-Pro v1Unspecified87.6%Vals AI
Vals AI · Tax Agent Bench v1: Weighted PassUnspecified45.1%Vals AI
Vals AI · Terminal-Bench 2.1Unspecified53.6%Vals AI
Vals AI · Vibe Code Bench v1.1Unspecified37.9%Vals AI
𝜏²-Bench TelecomNon-reasoning93.9%Artificial Analysis
𝜏²-Bench TelecomReasoning95.9%Artificial Analysis
𝜏³-BankingReasoning23.3%Artificial Analysis
Benchmark scores & sources

Original results, evaluation harnesses, and evidence behind this model.

BenchmarkBucketScoreHarnessEvidenceSource-recorded date
AA-LCR · source release snapshot; version not specifiedLong context70.2%
Reported settings & source

NVIDIA evaluation harness/settings per benchmark in model card; comparator scores are NVIDIA-reported under that protocol.

First-party reported result. Original cell: 70.2

nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 · Performance table: AA-LCR · reviewed 2026-09-06

Published configuration

Effort: Not specified

contributes to capability
lab self-reportReviewed 2026-09-06
Agents' Last ExamSupporting evidence9.2%Agents' Last Exam · OpenClaw

Effort: Not specified

official board

Supporting evidence outside the reviewed capability core

2026-10-20
AIME 2026Hard reasoning96.4%
Reported settings & source

Kimi K2.6 card, AIME 2026. Thinking mode; temperature 1.0, top-p 1.0, context 262144. Comparator settings: GPT-5.4 xhigh, Opus4.6 max, Gemini3.1Pro high. See benchmark-specific footnotes. Only own K2.6 and starred re-evaluations can form new comparisons. Reasoning: 98304 generation tokens; HLE full set.

Source SHA256 95db3be1d0473e482c7aa901f237ad7af8bb7929625a777b27428081bb8ea9f7. Own measurement or explicitly starred same-condition re-evaluation. Assess benchmark-specific protocol before admission.

Kimi K2.6 official model card · Evaluation Results table, AIME 2026, column 2 · reviewed 2026-11-22

Published configuration

Effort: Reasoning

lab self-report

No matched opponent in this evaluation unit

Reviewed 2026-11-22
AIME 2026 · source release snapshot; version not specifiedHard reasoning96.4%
Reported settings & source

MAI avg 4 runs, temperature 1, top-p .97; simple ReAct bash/string-replace; Terminal-Bench ignores predefined timeouts. Comparator configurations from cited reports.

First-party reported result; comparator results retain the source evaluation setup.

MAI-Thinking-1 technical report · Table 11, PDF page 53 · reviewed 2026-09-06

Published configuration

Effort: Not specified

contributes to capability
lab self-reportReviewed 2026-09-06
APEX-AgentsAgentic27.9%
Reported settings & source

Kimi K2.6 card, APEX-Agents. Thinking mode; temperature 1.0, top-p 1.0, context 262144. Comparator settings: GPT-5.4 xhigh, Opus4.6 max, Gemini3.1Pro high. See benchmark-specific footnotes. Only own K2.6 and starred re-evaluations can form new comparisons.

Source SHA256 95db3be1d0473e482c7aa901f237ad7af8bb7929625a777b27428081bb8ea9f7. Own measurement or explicitly starred same-condition re-evaluation. Assess benchmark-specific protocol before admission.

Kimi K2.6 official model card · Evaluation Results table, APEX-Agents, column 2 · reviewed 2026-11-22

Published configuration

Effort: Reasoning

lab self-report

No matched opponent in this evaluation unit

Reviewed 2026-11-22
Apex-Shortlist (no tools) · source release snapshot; version not specifiedSupporting evidence77.4%
Reported settings & source

NVIDIA evaluation harness/settings per benchmark in model card; comparator scores are NVIDIA-reported under that protocol.

First-party reported result. Original cell: 77.4

nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 · Performance table: Apex-Shortlist (no tools) · reviewed 2026-09-06

Published configuration

Effort: Not specified

lab self-report

Supporting evidence outside the reviewed capability core

Reviewed 2026-09-06
Apex-Shortlist (with tools) · source release snapshot; version not specifiedSupporting evidence73.2%
Reported settings & source

NVIDIA evaluation harness/settings per benchmark in model card; comparator scores are NVIDIA-reported under that protocol.

First-party reported result. Original cell: 73.2

nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 · Performance table: Apex-Shortlist (with tools) · reviewed 2026-09-06

Published configuration

Effort: Not specified

lab self-report

Supporting evidence outside the reviewed capability core

Reviewed 2026-09-06
BabyVisionSupporting evidence39.8%
Reported settings & source

Kimi K2.6 card, BabyVision. Thinking mode; temperature 1.0, top-p 1.0, context 262144. Comparator settings: GPT-5.4 xhigh, Opus4.6 max, Gemini3.1Pro high. See benchmark-specific footnotes. Only own K2.6 and starred re-evaluations can form new comparisons. Vision: 98304 tokens, average of three runs; Python condition uses 65536 tokens per step, at most50steps.

Source SHA256 95db3be1d0473e482c7aa901f237ad7af8bb7929625a777b27428081bb8ea9f7. Own measurement or explicitly starred same-condition re-evaluation. Assess benchmark-specific protocol before admission.

Kimi K2.6 official model card · Evaluation Results table, BabyVision, column 2 · reviewed 2026-11-22

Published configuration

Effort: Reasoning

lab self-report

Supporting evidence outside the reviewed capability core

Reviewed 2026-11-22
BabyVision (w/ python)Supporting evidence68.5%
Reported settings & source

Kimi K2.6 card, BabyVision (w/ python). Thinking mode; temperature 1.0, top-p 1.0, context 262144. Comparator settings: GPT-5.4 xhigh, Opus4.6 max, Gemini3.1Pro high. See benchmark-specific footnotes. Only own K2.6 and starred re-evaluations can form new comparisons. Vision: 98304 tokens, average of three runs; Python condition uses 65536 tokens per step, at most50steps.

Source SHA256 95db3be1d0473e482c7aa901f237ad7af8bb7929625a777b27428081bb8ea9f7. Own measurement or explicitly starred same-condition re-evaluation. Assess benchmark-specific protocol before admission.

Kimi K2.6 official model card · Evaluation Results table, BabyVision (w/ python), column 2 · reviewed 2026-11-22

Published configuration

Effort: Reasoning

lab self-report

Supporting evidence outside the reviewed capability core

Reviewed 2026-11-22
BrowseCompAgentic83.2%
Reported settings & source

Kimi K2.6 card, BrowseComp. Thinking mode; temperature 1.0, top-p 1.0, context 262144. Comparator settings: GPT-5.4 xhigh, Opus4.6 max, Gemini3.1Pro high. See benchmark-specific footnotes. Only own K2.6 and starred re-evaluations can form new comparisons.

Source SHA256 95db3be1d0473e482c7aa901f237ad7af8bb7929625a777b27428081bb8ea9f7. Own measurement or explicitly starred same-condition re-evaluation. Assess benchmark-specific protocol before admission.

Kimi K2.6 official model card · Evaluation Results table, BrowseComp, column 2 · reviewed 2026-11-22

Published configuration

Effort: Reasoning

lab self-report

No matched opponent in this evaluation unit

Reviewed 2026-11-22
BrowseComp · source release snapshot; version not specifiedAgentic83.2%
Reported settings & source

MiniMax launch figure methodology. SWE internal ClaudeCode 4 runs; Terminal 8CPU16GB2h128k Terminus2; Paper/PostTrain Ralph-loop12h; GDPval internal rubric grader; BrowseComp discard64k; OSWorld361tasks. Comparator provider/leaderboard scores where footnoted.

First-party reported result; comparator results retain the source evaluation setup.

MiniMax M3 model card benchmark figure · figures/benchmark.jpeg · reviewed 2026-09-06

Published configuration

Effort: Not specified

contributes to capability
lab self-reportReviewed 2026-09-06
BrowseComp · source release snapshot; version not specifiedAgentic61.3%
Reported settings & source

NVIDIA evaluation harness/settings per benchmark in model card; comparator scores are NVIDIA-reported under that protocol.

First-party reported result. Original cell: 61.3

nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 · Performance table: BrowseComp · reviewed 2026-09-06

Published configuration

Effort: Not specified

contributes to capability
lab self-reportReviewed 2026-09-06
BrowseComp (Agent Swarm)Agentic86.3%
Reported settings & source

Kimi K2.6 card, BrowseComp (Agent Swarm). Thinking mode; temperature 1.0, top-p 1.0, context 262144. Comparator settings: GPT-5.4 xhigh, Opus4.6 max, Gemini3.1Pro high. See benchmark-specific footnotes. Only own K2.6 and starred re-evaluations can form new comparisons.

Source SHA256 95db3be1d0473e482c7aa901f237ad7af8bb7929625a777b27428081bb8ea9f7. Own measurement or explicitly starred same-condition re-evaluation. Assess benchmark-specific protocol before admission.

Comparison limit: Agent swarm execution is not an individual model configuration.

Kimi K2.6 official model card · Evaluation Results table, BrowseComp (Agent Swarm), column 2 · reviewed 2026-11-22

Published configuration

Effort: Reasoning

lab self-report

Agent swarm execution is not an individual model configuration.

Reviewed 2026-11-22
CharXiv (RQ)Multimodal80.4%
Reported settings & source

Kimi K2.6 card, CharXiv (RQ). Thinking mode; temperature 1.0, top-p 1.0, context 262144. Comparator settings: GPT-5.4 xhigh, Opus4.6 max, Gemini3.1Pro high. See benchmark-specific footnotes. Only own K2.6 and starred re-evaluations can form new comparisons. Vision: 98304 tokens, average of three runs; Python condition uses 65536 tokens per step, at most50steps.

Source SHA256 95db3be1d0473e482c7aa901f237ad7af8bb7929625a777b27428081bb8ea9f7. Own measurement or explicitly starred same-condition re-evaluation. Assess benchmark-specific protocol before admission.

Kimi K2.6 official model card · Evaluation Results table, CharXiv (RQ), column 2 · reviewed 2026-11-22

Published configuration

Effort: Reasoning

contributes to capability
lab self-reportReviewed 2026-11-22
CharXiv (RQ) (w/ python)Supporting evidence86.7%
Reported settings & source

Kimi K2.6 card, CharXiv (RQ) (w/ python). Thinking mode; temperature 1.0, top-p 1.0, context 262144. Comparator settings: GPT-5.4 xhigh, Opus4.6 max, Gemini3.1Pro high. See benchmark-specific footnotes. Only own K2.6 and starred re-evaluations can form new comparisons. Vision: 98304 tokens, average of three runs; Python condition uses 65536 tokens per step, at most50steps.

Source SHA256 95db3be1d0473e482c7aa901f237ad7af8bb7929625a777b27428081bb8ea9f7. Own measurement or explicitly starred same-condition re-evaluation. Assess benchmark-specific protocol before admission.

Kimi K2.6 official model card · Evaluation Results table, CharXiv (RQ) (w/ python), column 2 · reviewed 2026-11-22

Published configuration

Effort: Reasoning

lab self-report

Supporting evidence outside the reviewed capability core

Reviewed 2026-11-22
Claw Eval (pass@3)Supporting evidence80.9%
Reported settings & source

Kimi K2.6 card, Claw Eval (pass@3). Thinking mode; temperature 1.0, top-p 1.0, context 262144. Comparator settings: GPT-5.4 xhigh, Opus4.6 max, Gemini3.1Pro high. See benchmark-specific footnotes. Only own K2.6 and starred re-evaluations can form new comparisons.

Source SHA256 95db3be1d0473e482c7aa901f237ad7af8bb7929625a777b27428081bb8ea9f7. Own measurement or explicitly starred same-condition re-evaluation. Assess benchmark-specific protocol before admission.

Kimi K2.6 official model card · Evaluation Results table, Claw Eval (pass@3), column 2 · reviewed 2026-11-22

Published configuration

Effort: Reasoning

lab self-report

Supporting evidence outside the reviewed capability core

Reviewed 2026-11-22
Claw Eval (pass^3)Supporting evidence62.3%
Reported settings & source

Kimi K2.6 card, Claw Eval (pass^3). Thinking mode; temperature 1.0, top-p 1.0, context 262144. Comparator settings: GPT-5.4 xhigh, Opus4.6 max, Gemini3.1Pro high. See benchmark-specific footnotes. Only own K2.6 and starred re-evaluations can form new comparisons.

Source SHA256 95db3be1d0473e482c7aa901f237ad7af8bb7929625a777b27428081bb8ea9f7. Own measurement or explicitly starred same-condition re-evaluation. Assess benchmark-specific protocol before admission.

Kimi K2.6 official model card · Evaluation Results table, Claw Eval (pass^3), column 2 · reviewed 2026-11-22

Published configuration

Effort: Reasoning

lab self-report

Supporting evidence outside the reviewed capability core

Reviewed 2026-11-22
Claw-Eval · source release snapshot; version not specifiedSupporting evidence61.5%
Reported settings & source

MiniMax launch figure methodology. SWE internal ClaudeCode 4 runs; Terminal 8CPU16GB2h128k Terminus2; Paper/PostTrain Ralph-loop12h; GDPval internal rubric grader; BrowseComp discard64k; OSWorld361tasks. Comparator provider/leaderboard scores where footnoted.

First-party reported result; comparator results retain the source evaluation setup.

MiniMax M3 model card benchmark figure · figures/benchmark.jpeg · reviewed 2026-09-06

Published configuration

Effort: Not specified

lab self-report

Supporting evidence outside the reviewed capability core

Reviewed 2026-09-06
CritPt (no tools) · source release snapshot; version not specifiedSupporting evidence9.1%
Reported settings & source

NVIDIA evaluation harness/settings per benchmark in model card; comparator scores are NVIDIA-reported under that protocol.

First-party reported result. Original cell: 9.1

nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 · Performance table: CritPt (no tools) · reviewed 2026-09-06

Published configuration

Effort: Not specified

lab self-report

Supporting evidence outside the reviewed capability core

Reviewed 2026-09-06
DeepSearchQA (accuracy)Supporting evidence83%
Reported settings & source

Kimi K2.6 card, DeepSearchQA (accuracy). Thinking mode; temperature 1.0, top-p 1.0, context 262144. Comparator settings: GPT-5.4 xhigh, Opus4.6 max, Gemini3.1Pro high. See benchmark-specific footnotes. Only own K2.6 and starred re-evaluations can form new comparisons.

Source SHA256 95db3be1d0473e482c7aa901f237ad7af8bb7929625a777b27428081bb8ea9f7. Own measurement or explicitly starred same-condition re-evaluation. Assess benchmark-specific protocol before admission.

Kimi K2.6 official model card · Evaluation Results table, DeepSearchQA (accuracy), column 2 · reviewed 2026-11-22

Published configuration

Effort: Reasoning

lab self-report

Supporting evidence outside the reviewed capability core

Reviewed 2026-11-22
DeepSearchQA (f1-score)Supporting evidence92.5%
Reported settings & source

Kimi K2.6 card, DeepSearchQA (f1-score). Thinking mode; temperature 1.0, top-p 1.0, context 262144. Comparator settings: GPT-5.4 xhigh, Opus4.6 max, Gemini3.1Pro high. See benchmark-specific footnotes. Only own K2.6 and starred re-evaluations can form new comparisons.

Source SHA256 95db3be1d0473e482c7aa901f237ad7af8bb7929625a777b27428081bb8ea9f7. Own measurement or explicitly starred same-condition re-evaluation. Assess benchmark-specific protocol before admission.

Kimi K2.6 official model card · Evaluation Results table, DeepSearchQA (f1-score), column 2 · reviewed 2026-11-22

Published configuration

Effort: Reasoning

lab self-report

Supporting evidence outside the reviewed capability core

Reviewed 2026-11-22
GDPVal · source release snapshot; version not specifiedAgentic50.4%
Reported settings & source

NVIDIA evaluation harness/settings per benchmark in model card; comparator scores are NVIDIA-reported under that protocol.

First-party reported result. Original cell: 50.4

nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 · Performance table: GDPVal · reviewed 2026-09-06

Published configuration

Effort: Not specified

contributes to capability
lab self-reportReviewed 2026-09-06
GDPval rubrics · source release snapshot; version not specifiedAgentic65.1%
Reported settings & source

MiniMax launch figure methodology. SWE internal ClaudeCode 4 runs; Terminal 8CPU16GB2h128k Terminus2; Paper/PostTrain Ralph-loop12h; GDPval internal rubric grader; BrowseComp discard64k; OSWorld361tasks. Comparator provider/leaderboard scores where footnoted.

First-party reported result; comparator results retain the source evaluation setup.

MiniMax M3 model card benchmark figure · figures/benchmark.jpeg · reviewed 2026-09-06

Published configuration

Effort: Not specified

contributes to capability
lab self-reportReviewed 2026-09-06
GPQA (no tools) · source release snapshot; version not specifiedHard reasoning91%
Reported settings & source

NVIDIA evaluation harness/settings per benchmark in model card; comparator scores are NVIDIA-reported under that protocol.

First-party reported result. Original cell: 91.0

nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 · Performance table: GPQA (no tools) · reviewed 2026-09-06

Published configuration

Effort: Not specified

contributes to capability
lab self-reportReviewed 2026-09-06
GPQA Diamond · source release snapshot; version not specifiedHard reasoning90.5%
Reported settings & source

MAI avg 4 runs, temperature 1, top-p .97; simple ReAct bash/string-replace; Terminal-Bench ignores predefined timeouts. Comparator configurations from cited reports.

First-party reported result; comparator results retain the source evaluation setup.

MAI-Thinking-1 technical report · Table 11, PDF page 53 · reviewed 2026-09-06

Published configuration

Effort: Not specified

contributes to capability
lab self-reportReviewed 2026-09-06